Comparisons between phenotypic master regulators in adipose.

(A) Clustered heatmap of Pearson correlations between the expression profiles for putative phenotypic master regulators in the TwinsUK adipose samples. The black bars to the right and bottom of the heatmap indicate for which phenotype each regulator was identified as a candidate master regulator. (B) Same as A, but between the activity profiles for putative phenotypic master regulators in the TwinsUK adipose samples. (C) Same as A, but between the expression profiles (for regulators indicated by row)and activity profiles (for regulators indicated by column) of putative phenotypic master regulators in the TwinsUK adipose samples. Note that the correlation coefficients for the heatmap in S7C Fig are not symmetrical about the diagonal since correlation between the expression profile of a row regulator and the activity profile of a column regulator is not the same as the converse, except along the diagonal (where the row and column regulators are identical). Consequently, though the rows and columns are in the same order, the clustering is based solely on columns (i.e. the vectors of correlation coefficients for the column regulators’ activity profiles versus the expression profiles for the row regulators). (D) Density plot of Pearson correlation between matched expression and activity for master regulators. (E) Density plot of Pearson correlation between unmatched expression and activity for master regulators. (F) Venn diagram demonstrating the number of master regulators that are unique or in common among the phenotypes analyzed.

(PDF)